The purpose of this study was to use the meta-analytic approach to examine the effects of caffeine ingestion on ratings of perceived exertion (RPE). Twenty-one studies with 109 effect sizes (ESs) met the inclusion criteria. Coding incorporated RPE scores obtained both during constant load exercise (n=89) and upon termination of exhausting exercise (n=20). In addition, when reported, the exercise performance ES was also computed (n=16). In comparison to placebo, caffeine reduced RPE during exercise by 5.6% (95% CI (confidence interval), −4.5% to −6.7%), with an equivalent RPE ES of −0.47 (95% CI, −0.35 to −0.59). These values were significantly greater (P<0.05) than RPE obtained at the end of exercise (RPE % change, 0.01%; 95% CI, −1.9 to 2.0%; RPE ES, 0.00, 95% CI, −0.17 to 0.17). In addition, caffeine improved exercise performance by 11.2% (95% CI; 4.6–17.8%). Regression analysis revealed that RPE obtained during exercise could account for ∼29% of the variance in the improvement in exercise performance. The results demonstrate that caffeine reduces RPE during exercise and this may partly explain the subsequent ergogenic effects of caffeine on performance.